It's no wonder you're confused about neutrality since you have never heard of the Canadian National Research Council work on acoustics in the 70s and 80s that influenced the entire audio industry, spawning renowned Canadian speaker makers like Paradigm and Energy. Google still exists, you know. So maybe you shouldn't go around calling people "absurd" and talking about "speaking volumes about your technical expertise" when the ignorance actually is coming from your side.
It's not a crude definition at all. It's actually a very sophisticated definition based on the acoustical research of the Canadian National Research Council in the 70s and 80s, which first quantified the elements of good sound related to frequency, dispersion, low distortion, etc.
But regardless of any semantics or terminology, the reported measurement results of xnor prove that +4 across the board increases volume over what is likely true zero, which potentially therefore also increases distortion, and regardless of distortion is not an accurate representation of the source material anyway. It therefore should be obvious to any so-called (or self-declared) "audiophile" that from the standpoint of true neutrality, +4 is to that extent undesirable.
«True zero» is a useless hypothetical term with no value. And there's no reason for the signal to have higher distortion from higher gain in the digital domain. It's rather the opposite, in that the lowered gain for EQ headroom might come with lower bit depth, but this scenario doesn't necessarily apply.
Clip+'s Normal EQ isn't flat...
...is definitely proven wrong by the measurements. I think there's nothing more to add.
True zero has a definite value as a term in this situation, because sansa's "Normal" setting is not at true zero, while a lot of people are being confused into thinking it is.
The measurements show that Sansa's "Normal" is an equivalent volume to +4, which likely adds distortion over true zero, and which therefore isn't flat. A distorted signal isn't flat.
A signal isn't "flat" just because all the frequencies are set to the same value. Is +99 flat, too? Not quite. There will be all kinds of distortion in that case way beyond what the frequency graph looks like for true zero flat. So it's not the same as zero, and therefore it's not really flat. The measurements prove you are wrong.